Given the confusion around this, let me try to enumerate all paths
(I will enumerate the end result as WG, but please substitute it with
"Stop" for the failure cases)
Idea --> WG
Idea --> SG --> WG
Idea --> BoF-1 --> WG
Idea --> BoF-1 --> BoF-2 --> WG
Idea --> BoF-1 --> SG --> WG
Idea --> BoF-1 --> BoF-2 --> SG --> WG
I am not too keen on SG --> BoF; but, if the sponsoring AD wants to do
it, I guess it may make sense to not explicitly disallow it. We can
probably look for that data during the experiment. If the SG failed to
develop a consensus charter during its life, a BoF session most likely
won't result in a different result (although one could argue that due to
time constraints or other reasons, the community evaluation of the
charter they developed can be done at a f2f meeting as opposed to at a
meeting; a similar argument has been made here by folks). To not flout
the 2418 rule, if two BoFs were already held and an SG was created, it's
WG or nothing after the SG.
So, we may also accommodate
Idea --> SG --> BoF-1 --> WG
Idea --> BoF-1 --> SG --> BoF-2 --> WG
The following two options do not make sense:
Idea --> SG --> BoF-1 --> BoF-2
Idea --> BoF-1 --> SG --> BoF-2 --> BoF (2418 disallows this)
Idea --> BoF-1 --> BoF-2 --> SG --> BoF (2418 disallows this)
(Hope I didn't make a mistake there ;) )
regards,
Lakshminath
On 10/10/2007 8:08 PM, Spencer Dawkins wrote:
Hi,
The complete text of the strawman -03 document is available here:
http://www.drizzle.com/~aboba/IAB/draft-aboba-sg-experiemnt-03.txt
Easily corrected typo, but just for ease of clicking, the correct URL is
http://www.drizzle.com/~aboba/IAB/draft-aboba-sg-experiment-03.txt
Now, for something completely different...
Since this is an experiment - please don't hold up the experiment while
you try to legalese-craft all the corner cases, but you might include a
"things to clarify if the experiment succeeds" list... which might
include this question...
From way down at the bottom of the 03 Introduction (which is really long,
but):
This document describes an RFC 3933 [RFC3933] experiment in the
Working Group formation process, known as the Study Group. Study
Groups MAY be formed by the IESG when there is evidence of clear
interest in a topic on the part of IETF participants and end-users,
and relevance to the Internet community has been demonstrated, but
other RFC 2418 [RFC2418] criteria relating to Working Group formation
(including creation of a satisfactory Charter) have not yet been met
as the result of a first or second Birds-of-a-Feather (BOF) session.
So far, so good... I'm not confused until the next paragraph:
Study Group milestones are focused on completion of prerequisites for
Working Group formation, and as a result they are expected to
conclude within a short time frame, with limited opportunities for
milestone extension. This Study Group experiment does not alter the
Working Group formation guidelines described in RFC 2418 [RFC2418]
Section 2.1, the processes relating to BoFs [BOF] or the Internet
Standards Process described in RFC 2026 [RFC2026].
Is everyone but me totally clear on how BOFs interact with SGs and WGs?
The way I'm reading this, the mainline path through this procedure is
that some community of interest requests a BOF, with a request that's
plausible enough for an AD to go for it, and then the IESG suggests a SG
after the BOF.
If the SG "succeeds", is there any opportunity to hold a second BOF
(which seems reasonable, as a WG-forming BOF that would have more
IETF-wide visibility), or does the SG have to go straight to WG (which
is the way I read version 03)?
And, for extra credit, if the answer is "yes", what if the community
held two BOFs before the SG formed (which would be the 2418 limit)? The
answer based on "does not alter the process" seems to be "no" - is that
what we want?
And a couple of nits...
Nit: there are places in this draft that say "charter", but since both
SGs and WGs have charters, it would be great if these occurences were
qualified as "SG charter" or "WG charter".
Nit: [BOF] is great advice (I've reviewed it at least once for Thomas),
but it's not "the processes relating to BOFs [BOF]". The last time I saw
[BOF], it was intended to be informational - the process is still
normatively described in 2418.
Thanks,
Spencer
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf